From owner-freebsd-net Tue Apr 25 17:13: 8 2000 Delivered-To: freebsd-net@freebsd.org Received: from khavrinen.lcs.mit.edu (khavrinen.lcs.mit.edu [18.24.4.193]) by hub.freebsd.org (Postfix) with ESMTP id 9245B37B534 for ; Tue, 25 Apr 2000 17:13:04 -0700 (PDT) (envelope-from wollman@khavrinen.lcs.mit.edu) Received: (from wollman@localhost) by khavrinen.lcs.mit.edu (8.9.3/8.9.3) id UAA41341; Tue, 25 Apr 2000 20:12:57 -0400 (EDT) (envelope-from wollman) Date: Tue, 25 Apr 2000 20:12:57 -0400 (EDT) From: Garrett Wollman Message-Id: <200004260012.UAA41341@khavrinen.lcs.mit.edu> To: Archie Cobbs Cc: csg@waterspout.com (C. Stephen Gunn), freebsd-net@FreeBSD.ORG Subject: Re: Proposal for ethernet, bridging, netgraph In-Reply-To: <200004260003.RAA55104@bubba.whistle.com> References: <20000425164027.A2553@waterspout.com> <200004260003.RAA55104@bubba.whistle.com> Sender: owner-freebsd-net@FreeBSD.ORG Precedence: bulk X-Loop: FreeBSD.org < said: > Well, the mbuf is aligned.. so if the IP header starts at mbuf + > 14, then it's not aligned... and m_adj() won't help either.. but > I don't know whether this is still valid though. However, a couple > of drivers take pains to handle this. It's vitally important for fast networking -- not just for NFS -- to have the IP header naturally aligned (as appropriate for the CPU). Unfortunately, some NICs (like the DEC Tulip) have broken DMA engines that can't start a transfer on an unaligned boundary, so those drivers end up having to copy some amount of data. (On some processors, it's faster to copy the whole packet than to take the unaligned accesses.) -GAWollman -- Garrett A. Wollman | O Siem / We are all family / O Siem / We're all the same wollman@lcs.mit.edu | O Siem / The fires of freedom Opinions not those of| Dance in the burning flame MIT, LCS, CRS, or NSA| - Susan Aglukark and Chad Irschick To Unsubscribe: send mail to majordomo@FreeBSD.org with "unsubscribe freebsd-net" in the body of the message